Presented By O’Reilly and Intel AI
Put AI to work
Sep 4-5, 2018: Training
Sep 5-7, 2018: Tutorials & Conference
San Francisco, CA

Executive Briefing: Is there a Moore’s law (not hardware related) for artificial intelligence

Ben Vigoda (Gamalon)
4:00pm-4:40pm Friday, September 7, 2018
Average rating: **...
(2.50, 2 ratings)

What you'll learn

  • Understand the current limitations on machine learning, what causes them, new approaches to overcome these limitations, and the implications and new applications that become available as a result

Description

Is there a Moore’s law for artificial intelligence? Current machine learning approaches are hitting limits in terms of available labeled data and supercomputing capacity to learn from data. This is especially true in natural language understanding. A single 15-word sentence can express more than 10 billion different ideas, so it is essentially impossible to provide enough labeled examples to cover every idea that a system might encounter. We end up training systems to categorize sentences into just tens or hundreds of domains or intents (i.e. classes), which means our current natural language systems make us feel like we are talking to Tarzan.

Ben Vigoda explains why these limits are actually temporary, and are really just an artifact of the first implementations of deep learning. With new advances, we should expect machines to follow a trend curve where they grow rapidly in their expertise and capabilities without requiring supervised learning; performing assembly of large quantities of human knowledge and skills into a single system. Over the last five years, next generation machine learning has been funded by DARPA and pioneered at MIT, Stanford, and Berkeley and is now reaching the market. At Gamalon, we are working with leading Global 2000 enterprises to use this technology to help power customer experience and digital transformation. The system can more easily learn to understand a very large number and combinations of customer issues in order to aid customers across the customer journey. In deployment, it listens to what each customer is saying, converts their inquiry into structured form in a database, provides analytics and continued learning, while offering an immediate meaningful action to guide the customer to the next best action, and provides human follow-up when appropriate.

Photo of Ben Vigoda

Ben Vigoda

Gamalon

Benjamin Vigoda is the CEO of Gamalon Machine Intelligence. Previously, Ben was technical cofounder and CEO of Lyric Semiconductor, a startup that created the first integrated circuits and processor architectures for statistical machine learning and signal processing. The company was named one of the “50 most innovative companies” by Technology Review and was featured in the Wall Street Journal, New York Times, EE Times, Scientific American, Wired, and other media. Lyric was successfully acquired by Analog Devices, and Lyric’s products and technology are being deployed in leading smartphones and consumer electronics, medical devices, wireless base stations, and automobiles. Ben also cofounded Design That Matters, a not-for-profit that for the past decade has helped solve engineering and design problems in underserved communities and has saved thousands of infant lives by developing low-cost, easy-to-use medical technology such as infant incubators, UV therapy, pulse oximeters, and IV drip systems that have been fielded in 20 countries. He has won entrepreneurship competitions at MIT and Harvard and fellowships from Intel and the Kavli Foundation/National Academy of Sciences and has held research appointments at MIT, HP, Mitsubishi, and the Santa Fe Institute. Ben has authored over 120 patents and academic publications. He currently serves on the DARPA Information Science and Technology (ISAT) steering committee. Ben holds a PhD from MIT, where he developed circuits for implementing machine learning algorithms natively in hardware.